{"url":"/sota/formation-energy-on-qm9","task":{"name":"Formation Energy","url":"/task/formation-energy","note":null},"dataset":{"name":"QM9","url":"/dataset/qm9"},"category":"Miscellaneous","categories":["Miscellaneous"],"category_note":null,"description":"On the QM9 dataset the numbers reported in the table are the mean absolute error in eV on the target variable U0 divided by U0's chemical accuracy, which is equal to 0.043.","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["MAE"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"MAE":"lower"}},"counts":{"rows":18,"rows_with_code":14,"rows_with_paper_page":18,"rows_dated":18,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"TensorNet","metrics":{"MAE":"0.09"},"uses_additional_data":false,"paper_date":"2023-06-10","paper":"/paper/tensornet-cartesian-tensor-representations","paper_url":"https://arxiv.org/abs/2306.06482v2","paper_title":"TensorNet: Cartesian Tensor Representations for Efficient Learning of Molecular Potentials","code":"https://github.com/torchmd/torchmd-net","n_code_links":3,"syntology":{"n_ran":3,"n_unverified":4,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"Wigner Kernels","metrics":{"MAE":"0.100 ± 0.003"},"uses_additional_data":false,"paper_date":"2023-03-07","paper":"/paper/wigner-kernels-body-ordered-equivariant","paper_url":"https://arxiv.org/abs/2303.04124v1","paper_title":"Wigner kernels: body-ordered equivariant machine learning without a basis","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"PAMNet","metrics":{"MAE":"0.136"},"uses_additional_data":false,"paper_date":"2023-11-19","paper":"/paper/a-universal-framework-for-accurate-and-1","paper_url":"https://arxiv.org/abs/2311.11228v1","paper_title":"A Universal Framework for Accurate and Efficient Geometric Deep Learning of Molecular Systems","code":"https://github.com/XieResearchGroup/Physics-aware-Multiplex-GNN","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":4,"model":"MXMNet","metrics":{"MAE":"0.137"},"uses_additional_data":false,"paper_date":"2020-11-15","paper":"/paper/molecular-mechanics-driven-graph-neural","paper_url":"https://arxiv.org/abs/2011.07457v1","paper_title":"Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures","code":"https://github.com/zetayue/MXMNet","n_code_links":3,"syntology":null},{"rank_in_archive_order":5,"model":"HMGNN","metrics":{"MAE":"0.138"},"uses_additional_data":false,"paper_date":"2020-09-26","paper":"/paper/heterogeneous-molecular-graph-neural-networks","paper_url":"https://arxiv.org/abs/2009.12710v1","paper_title":"Heterogeneous Molecular Graph Neural Networks for Predicting Molecule Properties","code":"https://github.com/shuix007/HMGNN","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":6,"model":"PhysNet-ens5","metrics":{"MAE":"0.14"},"uses_additional_data":false,"paper_date":"2019-02-22","paper":"/paper/physnet-a-neural-network-for-predicting","paper_url":"http://arxiv.org/abs/1902.08408v2","paper_title":"PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges","code":"https://github.com/MMunibas/PhysNet","n_code_links":6,"syntology":{"n_ran":1,"n_unverified":12,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"DeepMoleNet","metrics":{"MAE":"0.141"},"uses_additional_data":false,"paper_date":"2020-06-30","paper":"/paper/transferable-multi-level-attention-neural","paper_url":"https://doi.org/10.26434/chemrxiv.12588170.v1","paper_title":"Transferable Multi-level Attention Neural Network for Accurate Prediction of Quantum Chemistry Properties via Multi-task Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"xGPR -- Gaussian process, graph convolution kernel","metrics":{"MAE":"0.167"},"uses_additional_data":false,"paper_date":"2023-02-07","paper":"/paper/scalable-gaussian-process-regression-enables","paper_url":"https://arxiv.org/abs/2302.03294v2","paper_title":"Linear-scaling kernels for protein sequences and small molecules outperform deep learning while providing uncertainty quantitation and improved interpretability","code":"https://github.com/jlparki/xgpr","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"DimeNet","metrics":{"MAE":"0.185"},"uses_additional_data":false,"paper_date":"2020-03-06","paper":"/paper/directional-message-passing-for-molecular-1","paper_url":"https://arxiv.org/abs/2003.03123v2","paper_title":"Directional Message Passing for Molecular Graphs","code":"https://github.com/dmlc/dgl/tree/master/examples/pytorch/dimenet","n_code_links":4,"syntology":{"n_ran":1,"n_unverified":9,"n_samples":10,"n_pointer_only_licence":3}},{"rank_in_archive_order":10,"model":"PhysNet","metrics":{"MAE":"0.19"},"uses_additional_data":false,"paper_date":"2019-02-22","paper":"/paper/physnet-a-neural-network-for-predicting","paper_url":"http://arxiv.org/abs/1902.08408v2","paper_title":"PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges","code":"https://github.com/MMunibas/PhysNet","n_code_links":6,"syntology":{"n_ran":1,"n_unverified":12,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"MEGNet-Full","metrics":{"MAE":"0.21"},"uses_additional_data":false,"paper_date":"2018-12-12","paper":"/paper/graph-networks-as-a-universal-machine","paper_url":"http://arxiv.org/abs/1812.05055v1","paper_title":"Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals","code":"https://github.com/materialsvirtuallab/megnet","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"SchNet-edge-update","metrics":{"MAE":"0.242"},"uses_additional_data":false,"paper_date":"2018-06-08","paper":"/paper/neural-message-passing-with-edge-updates-for","paper_url":"http://arxiv.org/abs/1806.03146v1","paper_title":"Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials","code":"https://github.com/nrel/m2p","n_code_links":5,"syntology":{"n_ran":4,"n_unverified":18,"n_samples":22,"n_pointer_only_licence":4}},{"rank_in_archive_order":13,"model":"HIP-NN","metrics":{"MAE":"0.256"},"uses_additional_data":false,"paper_date":"2017-09-29","paper":"/paper/hierarchical-modeling-of-molecular-energies","paper_url":"http://arxiv.org/abs/1710.00017v1","paper_title":"Hierarchical modeling of molecular energies using a deep neural network","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":14,"model":"MEGNet-simple","metrics":{"MAE":"0.28"},"uses_additional_data":false,"paper_date":"2018-12-12","paper":"/paper/graph-networks-as-a-universal-machine","paper_url":"http://arxiv.org/abs/1812.05055v1","paper_title":"Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals","code":"https://github.com/materialsvirtuallab/megnet","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"ALIGNN","metrics":{"MAE":"0.30"},"uses_additional_data":false,"paper_date":"2021-06-03","paper":"/paper/atomistic-line-graph-neural-network-for","paper_url":"https://arxiv.org/abs/2106.01829v3","paper_title":"Atomistic Line Graph Neural Network for Improved Materials Property Predictions","code":"https://github.com/usnistgov/alignn","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"SchNet","metrics":{"MAE":"0.314"},"uses_additional_data":false,"paper_date":"2018-06-08","paper":"/paper/neural-message-passing-with-edge-updates-for","paper_url":"http://arxiv.org/abs/1806.03146v1","paper_title":"Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials","code":"https://github.com/nrel/m2p","n_code_links":5,"syntology":{"n_ran":4,"n_unverified":18,"n_samples":22,"n_pointer_only_licence":4}},{"rank_in_archive_order":17,"model":"MPNN","metrics":{"MAE":"0.49"},"uses_additional_data":false,"paper_date":"2017-04-04","paper":"/paper/neural-message-passing-for-quantum-chemistry","paper_url":"http://arxiv.org/abs/1704.01212v2","paper_title":"Neural Message Passing for Quantum Chemistry","code":"https://github.com/Microsoft/gated-graph-neural-network-samples","n_code_links":20,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":1}},{"rank_in_archive_order":18,"model":"HDAD+KRR","metrics":{"MAE":"0.58"},"uses_additional_data":false,"paper_date":"2017-02-17","paper":"/paper/machine-learning-prediction-errors-better","paper_url":"https://arxiv.org/abs/1702.05532","paper_title":"Machine learning prediction errors better than DFT accuracy","code":null,"n_code_links":0,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":12,"rows_with_any_sample_ran":9,"distinct_papers_with_graph_line":9,"distinct_papers_with_any_sample_ran":7,"samples_over_distinct_papers":{"n_ran":18,"n_unverified":54,"n_samples":72,"n_pointer_only_licence":14,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":23,"n_unverified":93,"n_samples":116,"n_pointer_only_licence":18,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}